Hello,
In the past I've proposed to capture musical notes in Codec2 in a better
way; David's response was that this was still open, but I am now seeing
so much implementation work that makes it more and more involved to change
things. So I'm trying to "reopen" the issue, before it's too late.
Codec2 encodes the base frequency of the voice on a linear scale, with a
somewhat rough resolution. The scale will probably sound just as well
if it is a logarithmic scale, which also happens to match workings of
our hearing.
The experience of singing through a telephone would be dramatically
impacted by the choice of scale. If frequencies are rounded off to
the nearest match on a linear scale, then they are bound to be off-pitch.
On the other hand, if musical halftones (or perhaps halfway these) were
used as a rounded value, then Codec2 would pass musical tones at the
perfect pitch -- the roundoff could actually improve the song ;-)
Two subsequent musical halfnotes are a factor of 12'th root of 2 apart,
so 12 such halfnotes make up a factor 2 or an octave. The frequency
440 Hz is defined as an A. Using step sizes of the 24'th root of 2
or 1.0293022366434921 should give a comparable experience, but be
friendly to music. It is also slightly more compact.
Am I the only one who would find it a pitty if we missed out on this,
if the only reason would be the existing implementations in the field?
Thanks,
Rick van Rein
OpenFortress
------------------------------------------------------------------------------
Live Security Virtual Conference
Exclusive live event will cover all the ways today's security and
threat landscape has changed and how IT managers can respond. Discussions
will include endpoint security, mobile security and the latest in malware
threats. http://www.accelacomm.com/jaw/sfrnl04242012/114/50122263/
_______________________________________________
Freetel-codec2 mailing list
Freetel-codec2@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/freetel-codec2
